Wednesday, January 2, 2019

A comparison of 2V and 4V Ducati 750 engines

.
When I did my 2V cam profile measuring (here), I had the chance to test a theory that I had, being that the cam profiles for the 4V Strada cams - marked A1 - were simply the 2V profiles of the time - the F1 cam that became the R grind used in the 600 and 750 engines from 86 to 02.  Turns out I was probably half right.

The inlet profile looks mostly the same.  The exhaust has a fatter nose for want of a better description, and is more asymmetrical, biased to the retarded side.  I found it a bit odd that they would just use the 2V profile, but while in reality the 2V cam was lacking quite a bit of lift when used with the 41mm inlet valve in the 750 F1, its 9.5mm of lift was more than enough for a 32mm inlet valve (as the 748cc 4V initially had).  Lift required is based on the fact that a poppet valve flow area - if you like the curtain you get around the valve od when lifted - is equal to the valve head area when the lift is 1/4 of the valve head diameter.  So if you have 32mm valve you need 8mm lift.  And 41mm needs 10.25mm lift.  It's a rough guide sort of thing.  By the time they got to the production 748, the inlet valves had grown to 33mm and exhausts to 29mm.

(Thinking back to the above now, after having written much of what follows, I've come to the conclusion that this comparison is flawed from the start.  The 750F1 cam simply doesn't have enough lift for its valves.  Maybe someone on the 4V project at Ducati (Bordi and his team) saw the profile and thought "That's about right for a 32mm valve, I'll have that".  Fair play to them.)

(And another - the 85 750F1 has small 500 Pantah valve sizes, so maybe for them the valve lift was ok.)

To the profile comparison.  Blue and red are A1 inlet and exhaust, green and orange the R inlet and exhaust.



I then wondered if they'd used the P grind from the hot F1 series - Montjuich, Laguna Seca and Santamonica - for the A cam used in the 851 Tricolore Kit and SP series.  But clearly not, the P has much more lift and more duration.

Some time later it occured to me that this meant there were a couple of engines from the same manufacturer with the same bore and stroke using the same inlet cam profile and a similar exhaust cam profile in both 2V and 4V configurations.  Those two engines are the 750 2V and 748 4V, both with bore and stroke of 88 x 61.5mm.

As such, I thought a comparison of 2V and 4V was in order.

In basic engine design, the potential peak torque value is based predominantly on capacity.  There is a measure called Brake Mean Effective Pressure, which is basically torque per cc, and it's somewhat consistant with similar engine designs and a measure of efficiency, both volumetric (how much air the engine can suck in and then trap) and combustion (compression ratio, chamber design).  The actual outright potential efficiency in terms of energy in versus power out is related to the compression ratio, although measures to reduce losses are a big part of engine design these days.

In contrast, the potential peak power value is not capacity based, but air flow based.  Simply, how much air can the engine flow through itself.  This also overlaps with mechanical design and maximum rpm potential.  If you're designing an engine for racing, for instance, you'd start by defining the maximum rpm you're going to use, which is mostly bound by piston speed.  Maximum piston speed is not an enforced limit, but a number that is in reality a development boundary to be pushed and see what happens.  The old "to finish first, one must first finish" kind of thing.  Usually defined as "mean piston speed", which is the average speed over the journey from top to bottom.  The actual maximum piston speed, which occurs about mid stroke, also varies with connecting rod length.  Seems that 30m/s is about the current maximum "mean" used.

With the limits of capacity and maximum piston speed set, the stroke is defined and then the bore defined by that.  Or, in MotoGP for instance, where the capacity limit is 1000cc and bore limit 81mm, those limits give a minimum stroke of 48.5mm, which in turn, if the maximum rpm used is 17,000, gives a minimum mean piston speed of 27.5m/s.

This can also depend on physical properties of components used too, and the limit there, if not specified by rules, is generally cost.  The old "how fast do you want to spend".  As a guide, when Ducati were at the end of the 999 race life, they were revving them to 14,500rpm to try to be competitive.  With a 63.5mm stroke, that gives a mean piston speed of 30.7m/s.  Pistons and connecting rods were not limited by the rules, though, unlike the 1198 era when production parts had to be used.  Through material spec and cost, this effectively limits the maximum mean piston speed indirectly.  This, along with the physical limit of further enlargement of bore size on the Pantah derived 1198 crankcases, is what led to the Panigale engine development.  Bigger bore means shorter stroke, and the shorter stroke was what they needed at the time.

The main secondary variation in terms of possible peak torque production is that the 4V engine has more compression, which realistically is 1/ simply a by product of the required (and better) pent roof combustion chamber design, and 2/ also permitted by that same chamber design with its ability to better resist detonation - ie, a smaller, shallower chamber with valves at a much smaller included angle, large squish areas and a central spark plug.  It is also water cooled, again making it a better high performance engine.  So the generic 4V pent roof head configuration has combustion chamber design advantages you can't get in a 2V without restricting valve size markedly.  The 2V chamber in the 750 engine is a (largely) hemispherical type chamber, which tends to give a deep chamber with much more valve angle than the 4V design simply to get the required valves in.  If you want high compression, you use a piston with a dome that in itself causes issues with flame front travel and increases the potential for detonation.

In so many ways the shallower 4V chamber and flat top piston are just better.

748 chamber on the left, 750 on the right.  748 valve sizes are 33/29mm inlet/exhaust, the 750 is one of my bigger valve heads with 42.5/37mm.  Std 750 is 41/35mm.



Comparing power and torque outputs was the aim of this, so we'll get into it.  Blue is 748, red 750SS and green 750SSie.


The power graph shows the 748 peak power value around 50% more than the 750, at an rpm similarly 50% higher.  This makes sense, as power is defined as torque x rpm.  Same torque at 50% higher rpm will give 50% more power.  If you compare the valve sizes, the 748 has about 30% more valve area.


The torque output, on the other hand, shows some of the implications of trying to spread the rpm range over which the power is produced.  The 2V curves are both generally just rounded curves, whereas the 748 has various peaks and troughs.  This is usually what happens as you widen the rpm range you're tuning for.  At lower rpm, as a general rule, you want things to be longer.

The difference between the 750SS and 750SSie is carburettors vs fuel injection.  The 750 engine with the Mikuni carbs has long inlet manifolds that position the carbs side by side.  The 750 ie engine has much shorter manifolds and throttle bodies arranged at 90 to each other, giving the appearance of a much shorter inlet tract.  In reality, the carb model has short rubbers into the airbox, whereas the ie model has much longer rubber trumpets inside the airbox (much longer on the vertical cylinder) that would mean the overall trumpet to port dimension would not be that much different between the two models.  Keep in mind everything else engine wise - heads, valves, pistons, cams - are the exact same parts.

One thing that is done when you try to extend the power range is to make the inlet shorter.  This helps the overall peak power value, but while it also reduces the peak torque value, it will increase the average torque figure over the wider rpm range and introduce the peaks and troughs.  The narrower the rpm you tune for, the more specific you can be in your tuning setup - inlet length, etc, and that will lead to a higher peak torque figure.  Great for a unique application such as a stationary engine, not so good for general motorcycle use.

There is a great dyno graph that I have never been able to find online - the only place I can recall seeing it in Sir Harry Ricardo's book - from D type Jaguar testing that shows the power curve with inlet lengths varying from maybe 6 to 36 inches.  At 36" the peak torque is well above that of 6", but the rpm range is much narrower.  As an implications and compromises of tuning illustration, it's particularly simple, elegant and quite definitive.  I might have to go to a Uni library and photocopy it some day.

Another point here is that while the inlet cams in this instance are all spec'd at the same timing - 119 degrees inlet centreline, with the exhausts at 106 on the 2V and 112 on the 4V, in reality the 2V would be pretty close to that or even maybe 1 or 2 degrees advanced, whereas the 4V would have all cams retarded to some extent.  We used to find 4V inlets from 5 to 10 degrees retarded out of the box, and when we reset the timing to 108/108 we did sometimes move inlet cams over 20 degrees.  This gives a better shape to the 748 torque curve, as below.  Blue is as delivered, red 108/108.


Incidentally, when this cam first appeared in the 750F1, the timing was 10 degrees more advanced than in the 750 Paso, which makes much more sense to me.  I have read that the timing was changed for "emissions" reasons, which really makes no sense to me as there was no motorcycle emissions testing back then.  I'd speculate it was done for two reasons.

1/ The Paso, with the longer inlet manifolds and Weber carb, had a more midrange orientated power delivery, and retarding the cam timing gave a power delivery more in keeping with what was expected and helped increase the outright power figure.

2/ With the fully enclosed body of the Paso and a compression ratio much the same as the F1 (spec is 10:1 in the Paso manual), even with the oil cooling system I'd think the risk of detonation would have been higher so retarding the cam timing would reduce the midrange cylinder pressure.

By the time the 750SS appeared it had pistons which are the slightest bulge off flat, and the listed compression ratio of 9:1 meant that there was no detonation in sight.  The other difference with the later pistons was that the inlet valve reflief was shallower, so the most advanced I was game to run those cams was 114 degree inlet centreline.  Not sure if you could run 110 on a 750 Paso or not.

Comparing a 750SS with timing at 119 and 114 we get the following.  If we could have run 110 and picked up a little more at the 5500 rpm torque peak and lost a little more over 8000 rpm it would definitely impact the feel of the power delivery, and that may have been a bad thing when you only have 2V motors on offer and people who just love pointlessly revving the snot out of things.


Inevitably, the two previous graphs lead to this one, being 748 vs 750 with more consistant cam timing settings.


Peak torque is almost the same, peak power about the same difference as before.


Yeah, but.....

Frankly, there's more yebits to this story than story.  It was sort of a good idea that has, realistically, taken it in the neck quite a few times along the way.

One could argue that we could make the 750 engine a better one for this comparison by giving it a different cam with more appropriate lift.  Totally negating the comparison, of course.  And its under valved, so lets fix that too.  And the comp is too low compared to the 748.  Luckily, I prepared one earlier.  My 750 engine with 900 cams, 42.5/37 valves and Ferracci 12:1 pistons.  The 900 cams have 12 degrees more duration and 2.5mm more lift.


Even though the cam duration has gone up, the torque curve is still a single peak, typical of the 2V motors.  

But, if we're going to make the 750 a bit better, we can make the 748 better too.  Along the way they had a couple of goes at the 748.  The SP/SPS with its great big old school cams, and the R with the bigger throttles and valves and half big cams.  Neither made a better road bike - it always dismayed me when people who bought 748R came in and claimed their old 748 had more midrange.  Of course it did.  The concept of an homologation special is not understood by many.

748 in blue, 748SPS in green, 748R in red.  Different shape curves, or more importantly, different amounts of torque over 10,000rpm.  Where the power is.  But, the peak torque value is pretty consistant across the 3, as it has been for all the graphs so far.


But (I'm liking these buts), if we're going to improve the 750, we might as well improve the 748 too.  Properly.  That involves bigger valves again - 37/30.5, larger than the 748/748SP at 33/29 or the 748R's 36/30, but reduced cam duration - at least 20 degrees on the inlet and 16 on the exhaust with appropriate lift.  Let's call it a 749.

The totally different exhaust designs may influence the 749 output, but in comparison to the above 748's adding a base model 749 curve in burgundy shows how much better it is.  Again, 750cc so similar peak torque, but the shorter cam duration and bigger valves combine to give better low rpm and high rpm torque compared to the 748.  The valve of good heads and just enough cam.  Although, on the road, it's kind of a boring power delivery.  Having the torque rise more noticeably with the rpm does make them feel more racy.


Power curves show the 749 is much more linear.  I think it'd make a good naked bike motor.


I might stop there.  I've kind of lost the direction of this, which really is just me thinking about it too much.



5 comments:

  1. This comment has been removed by the author.

    ReplyDelete

  2. This may, or may not, be what you were referring to.
    http://www.emeraldm3d.com/articles/emr-adj-length-intake/

    ReplyDelete
  3. Awesome Brad.
    One thing you’re yet to compare though is the ports themselves. The hemi head ports are awful. Short-turns are terrible and ports often just too big as evidenced on flow benches. This is the true and defining difference between 2V and 4V elaborations in my opinion. The cam specs will never be properly realised in a OEM 2V inlet tract. This is my opinion on the subject.
    Cheers Brad

    ReplyDelete
    Replies
    1. I agree with Nick. The 2-valve heads had crappy ports. They were the victim of starting at 500cc and, also, packaging concerns with carbs.

      Delete